Proc. Natl. Acad. Sci. USA Vol. 88, pp. 10009-10013, November 1991 Cell Biology
Characterization of a ceramide-activated protein kinase: Stimulation by tumor necrosis factor a SHALINI MATHIAS, KENNETH A. DRESSLER,
AND
RICHARD N. KOLESNICK
Laboratory of Signal Transduction, Memorial Sloan-Kettering Cancer Center, New York, NY 10021
Communicated by Lloyd J. Old, August 12, 1991 (received for review June 3, 1991)
event in cellular activation (6-8). Further, a synthetic ceramide N-acetylsphingosine could replace these agents in induction of monocytic differentiation of these cells. This laboratory has also provided some direct evidence for second-messenger function for ceramide. Davis and coworkers (15-17) originally showed that sphingosine induced epidermal growth factor receptor (EGFR) phosphorylation on Thr-669 in A-431 human epidermoid carcinoma cells by a mechanism that did not involve protein kinase C. This laboratory demonstrated that sphingosine was rapidly converted to ceramide by these cells and that ceramide induced identical phosphorylation (18). These studies were interpreted as evidence that ceramide had bioeffector properties and might mediate, in part, the action of exogenous sphin-
ABSTRACT Recent investigations have identified a signaltransduction system involving sphingomyelin and derivatives. In this paradigm, sphingomyelin hydrolysis by a sphingomyelinase generates ceramide, which may be converted to the protein kinase C inhibitor sphingosine or to ceramide 1-phosphate. Ceramide may have second-messenger function because it induces epidermal growth factor receptor phosphorylation, presumably on Thr-669 in A-431 cells. The present studies describe a kinase that may mediate ceramide action. With a 19-amino acid epidermal growth factor receptor peptide containing Thr-669, a membrane-bound activity that phosphorylated the peptide was detected in A-431 cells. Activity was linearly related to ATP (0.3-300 ,uM) and peptide concentration (0.02-1 mg/ml), possessed a physiologic pH optimum (pH 7.0-7.4), and was Mg2+-dependent. Other cations-Ca ", Mn2+, and Zn2+-were ineffective. Natural and synthetic ceramide induced time- and concentration-dependent enhancement of kinase activity. Ceramide (0.5 ,uM) increased kinase activity 2-fold by 30 s, and activity remained elevated for at least 15 min. As little as 0.001 ,LM ceramide was effective, and 1 ,LM ceramide induced maximal phosphorylation. Sphingosine was similarly effective. Because tumor necrosis factor (TNF) a rapidly induces sphingomyelin hydrolysis to ceramide during monocytic differentiation of HL-60 cells, its effects on kinase activity were assessed. Kinase activity was increased 1.5-fold at 5 min and 2-fold at 2 hr in membranes derived from TNF-stimulated cells. The effective concentration range was 3 pM-30 nM TNF. Exogenous ceramide induced a similar effect. In sum, these studies demonstrate the existence of an unusual Mg2+-dependent ceramide-activated protein kinase that may mediate some aspects of TNF-a function.
gosine. The present studies were done to identify the kinase that mediated the effect of ceramide on EGFR phosphorylation. The substrate used was a synthetic peptide derived from the amino acid sequence around Thr-669 of the EGFR. These studies demonstrate that A-431 human epidermoid carcinoma cells and HL-60 cells contain a ceramide/sphingosineactivated protein kinase. Further, this kinase is stimulated by TNF-a, which elevates the cellular ceramide level and induces phosphorylation of several proteins (19-24), including the EGFR, as an early event in cellular activation. These studies provide initial evidence for a sphingolipid-activated, protein kinase-mediated signaling system.
EXPERIMENTAL PROCEDURES Materials. Ceramide (type III), sphingosine, palmitic acid, cholera toxin, hexamethylene bisacetamide, retinoic acid, butyrate, leupeptin, and buffers were from Sigma. Fetal bovine serum (FBS) was from GIBCO. [y-32P]ATP (3000 Ci/mmol; 1 Ci = 37 GBq) was from New England Nuclear. P81 phosphocellulose paper was from Whatman. Liquid scintillation solution (Liquiscint) was from National Diagnostics (Sommerville, NJ). HPLC grade solvents were from Fisher. The EGFR peptide (amino acids 663-681, NH2-Glu-
Recent investigations from this and other laboratories have identified a metabolic pathway involving sphingomyelin and derivatives that may be involved in signal transduction (1-8). This pathway is initiated by hydrolysis of sphingomyelin to ceramide via the action of a sphingomyelinase. Ceramide may then be deacylated to sphingoid bases, putative inhibitors of protein kinase C (9-12), or phosphorylated to the sphingolipid ceramide 1-phosphate by the action of a recently described calcium-dependent ceramide kinase (4, 5, 13). The biologic role of ceramide 1-phosphate and regulation of the kinase that mediates its synthesis have not yet been determined. This pathway appears specific for ceramide derived from sphingomyelin, as ceramide derived from glycosphingolipids is not converted either to sphingoid bases (14) or to ceramide 1-phosphate (4). Recently, Hannun and coworkers (6-8) have provided evidence that this sphingomyelin pathway may be involved in signal transduction. Tumor necrosis factor (TNF) a, y interferon, and 1,25-dihydroxyvitamin D3, factors that induce monocytic differentiation of HL-60 promyelocytic cells, all stimulate sphingomyelin degradation to ceramide as an early
Leu-Val-Glu-Pro-Leu-Thr-Pro-Ser-Gly-Glu-Ala-Pro-AsnGln-Ala-Leu-Leu-Arg-COOH) was synthesized by using an Applied Biosystems model 431A machine and purified by reverse-phase HPLC. C8-ceramide (N-octanoylsphingosine; C8-cer) and TNF-a were gifts of R. Bittman (Queens College, New York) and D. Donner (Sloan-Kettering Institute), respectively. Cell Culture. A-431 human epidermoid carcinoma cells were grown in monolayer culture in a 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) and Ham's F-12 medium containing 10% FBS and were harvested by trypsinization as described (18). HL-60 cells were grown in suspension culture in RPMI 1640 medium containing 10% FBS and supplements, as described (3). On the day of an
The publication costs of this article were defrayed in part by page charge
Abbreviations: FBS, fetal bovine serum; TNF, tumor necrosis factor; C8-cer, C8-ceramide or N-octanoylsphingosine; EGFR, epidermal growth factor receptor.
payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
10009
10010
Cell
Biology: Mathias et al.
experiment HL-60 cells were resuspended (1 x 106 cells per ml) in RPMI 1640 medium/1% FBS for 2 hr before stimulation with lipid activators and differentiating agents. Membrane Preparation. Cells (3 X 107/ml) were homogenized with a tight-fitting Dounce homogenizer at 40C in buffer (25 mM Hepes, pH 7.4/5 mM EGTA/50 mM NaF/leupeptin at 10 Ag/ml) as described by Countaway et al. (17). The homogenate was centrifuged at 500 X g for 5 min, and the postnuclear supernatant was centrifuged at 200,000 x g for 30 min. The microsomal membrane pellet was resuspended (7.5 pug of membrane protein per ul for A-431 cells and 1.5 Ag/,l~ for HL-60 cells) in homogenizing buffer. Membranes were prepared fresh daily. Assay of Kinase Activity. For most experiments, the reaction mixture contained 25 Al of microsomal membrane or postnuclear supernatant, 50 ,ul of EGFR peptide (4 mg/ml in 25 mM Hepes, pH 7.4) and 125 A.l of buffer (50 mM Hepes, pH 7.4/20 mM MgCl2) (17). Phosphorylation was initiated at 22°C by addition of 50 pl of [yy-32P]ATP (150 ,M final concentration; 4000 dpm/pmol). For studies with lipid activators, ceramide and other lipids were dried under N2 and resuspended in the kinase assay buffer by bath sonication for 2 min at 37°C. The reaction was terminated at the indicated times by addition of 50 pu of 0.5 M ATP in 90% formic acid. Unless otherwise indicated, all assays were done under conditions determined as linear for time and enzyme concentration. Enzyme activity was determined from the transfer of 32p from the y position of ATP to EGFR peptide and the specific radioactivity of [y-32P]ATP. Phosphorylated peptide was quantified by two separate methods. For initial studies, samples were spotted on phosphocellulose paper, washed in 1 M acetic acid/4 mM pyrophosphate and subjected to liquid scintillation counting, as described (25). Values obtained from a boiled blank or a sample lacking peptide were subtracted from each determination. Alternatively, HPLC was done as described (17). For these studies, samples were first applied to a C18 Sep-Pak cartridge and eluted with 99.9% acetonitrile/0.1% trifluoroacetic acid. The eluates were lyophilized, resuspended in 6 M guanidine hydrochloride/200 mM Tris, pH 8.5 and applied to a C18 reverse-phase column (Dynamax, 4.6 mm i.d., Rainin, Woburn, MA). The peptide was eluted with a linear gradient (1% per min) of acetonitrile at a flow rate of 1 ml/min and was detected by measuring the Cerenkov radiation associated with 1-ml fractions. Phosphoamino Acid Analysis. To determine which amino acid was phosphorylated, phosphoamino acid analysis of the peptide was done. The phosphopeptide peak obtained by HPLC was subjected to partial acid hydrolysis (1 hr at 110°C in 6 M HCI). The hydrolysates were dried, resuspended in 250 ,ul of water, and applied to a Dowex AG1-X8 column (BioRad). Amino acids were eluted with 0.5 M HCI, dried, and analyzed by thin-layer electrophoresis, as described by Hunter and coworkers (26). Individual phosphoamino acids were identified by ninhydrin staining of carrier phosphoamino acids and by autoradiography. Other Procedures. Protein was measured by the method of Bradford (27). Statistics. Statistical analysis was performed by t test and linear regression analysis by the method of least squares.
RESULTS Davis and coworkers (15, 17) showed that addition of sphingosine to A-431 cells enhanced phosphorylation of the EGFR on Thr-669. Subsequently, work from this laboratory showed that sphingosine was rapidly converted to ceramide in these cells and that exogenous ceramide induced identical effects (18). To investigate the kinase that mediated ceramide action,
Proc. Natl. Acad. Sci. USA 88 (1991) a synthetic peptide corresponding to the sequence around Thr-669 was used as substrate. Initial studies were done to determine the kinetics of phosphorylation ofthe EGFR peptide. The conditions for this assay were adapted from Davis and coworkers (17, 25). Postnuclear supernatant was used as a source of enzyme activity, and samples were spotted on phosphocellulose paper to measure phosphorylated peptide. The kinetics of phosphorylation appeared biphasic. Initial rapid incorporation of 32p into peptide for 10 min was followed by incorporation at a slightly reduced rate for as long as 30 min (Fig. 1). Subsequent studies were done to optimize the assay. Kinase activity was found by Lineweaver-Burke analysis to be linearly related (r = 0.98) to substrate concentration for ATP (0.3-300 ,M) and EGFR peptide (0.02-1 mg/ml) at 5 min of stimulation. Apparent Km values for ATP of 15 ,uM and for EGFR peptide of 0.25 mg/ml were derived. Apparent Vmax values ranged from 100-200 pmol-min-1/mg of protein-'. All subsequent studies were done with 150 ,M ATP and EGFR peptide at 4 mg/ml. Under these conditions, substrate concentration was not rate limiting. An additional set of studies assessed the pH optimum for the kinase activity. There was no measurable activity at pH values
Characterization of a ceramide-activated protein kinase: Stimulation by tumor necrosis factor a SHALINI MATHIAS, KENNETH A. DRESSLER,
AND
RICHARD N. KOLESNICK
Laboratory of Signal Transduction, Memorial Sloan-Kettering Cancer Center, New York, NY 10021
Communicated by Lloyd J. Old, August 12, 1991 (received for review June 3, 1991)
event in cellular activation (6-8). Further, a synthetic ceramide N-acetylsphingosine could replace these agents in induction of monocytic differentiation of these cells. This laboratory has also provided some direct evidence for second-messenger function for ceramide. Davis and coworkers (15-17) originally showed that sphingosine induced epidermal growth factor receptor (EGFR) phosphorylation on Thr-669 in A-431 human epidermoid carcinoma cells by a mechanism that did not involve protein kinase C. This laboratory demonstrated that sphingosine was rapidly converted to ceramide by these cells and that ceramide induced identical phosphorylation (18). These studies were interpreted as evidence that ceramide had bioeffector properties and might mediate, in part, the action of exogenous sphin-
ABSTRACT Recent investigations have identified a signaltransduction system involving sphingomyelin and derivatives. In this paradigm, sphingomyelin hydrolysis by a sphingomyelinase generates ceramide, which may be converted to the protein kinase C inhibitor sphingosine or to ceramide 1-phosphate. Ceramide may have second-messenger function because it induces epidermal growth factor receptor phosphorylation, presumably on Thr-669 in A-431 cells. The present studies describe a kinase that may mediate ceramide action. With a 19-amino acid epidermal growth factor receptor peptide containing Thr-669, a membrane-bound activity that phosphorylated the peptide was detected in A-431 cells. Activity was linearly related to ATP (0.3-300 ,uM) and peptide concentration (0.02-1 mg/ml), possessed a physiologic pH optimum (pH 7.0-7.4), and was Mg2+-dependent. Other cations-Ca ", Mn2+, and Zn2+-were ineffective. Natural and synthetic ceramide induced time- and concentration-dependent enhancement of kinase activity. Ceramide (0.5 ,uM) increased kinase activity 2-fold by 30 s, and activity remained elevated for at least 15 min. As little as 0.001 ,LM ceramide was effective, and 1 ,LM ceramide induced maximal phosphorylation. Sphingosine was similarly effective. Because tumor necrosis factor (TNF) a rapidly induces sphingomyelin hydrolysis to ceramide during monocytic differentiation of HL-60 cells, its effects on kinase activity were assessed. Kinase activity was increased 1.5-fold at 5 min and 2-fold at 2 hr in membranes derived from TNF-stimulated cells. The effective concentration range was 3 pM-30 nM TNF. Exogenous ceramide induced a similar effect. In sum, these studies demonstrate the existence of an unusual Mg2+-dependent ceramide-activated protein kinase that may mediate some aspects of TNF-a function.
gosine. The present studies were done to identify the kinase that mediated the effect of ceramide on EGFR phosphorylation. The substrate used was a synthetic peptide derived from the amino acid sequence around Thr-669 of the EGFR. These studies demonstrate that A-431 human epidermoid carcinoma cells and HL-60 cells contain a ceramide/sphingosineactivated protein kinase. Further, this kinase is stimulated by TNF-a, which elevates the cellular ceramide level and induces phosphorylation of several proteins (19-24), including the EGFR, as an early event in cellular activation. These studies provide initial evidence for a sphingolipid-activated, protein kinase-mediated signaling system.
EXPERIMENTAL PROCEDURES Materials. Ceramide (type III), sphingosine, palmitic acid, cholera toxin, hexamethylene bisacetamide, retinoic acid, butyrate, leupeptin, and buffers were from Sigma. Fetal bovine serum (FBS) was from GIBCO. [y-32P]ATP (3000 Ci/mmol; 1 Ci = 37 GBq) was from New England Nuclear. P81 phosphocellulose paper was from Whatman. Liquid scintillation solution (Liquiscint) was from National Diagnostics (Sommerville, NJ). HPLC grade solvents were from Fisher. The EGFR peptide (amino acids 663-681, NH2-Glu-
Recent investigations from this and other laboratories have identified a metabolic pathway involving sphingomyelin and derivatives that may be involved in signal transduction (1-8). This pathway is initiated by hydrolysis of sphingomyelin to ceramide via the action of a sphingomyelinase. Ceramide may then be deacylated to sphingoid bases, putative inhibitors of protein kinase C (9-12), or phosphorylated to the sphingolipid ceramide 1-phosphate by the action of a recently described calcium-dependent ceramide kinase (4, 5, 13). The biologic role of ceramide 1-phosphate and regulation of the kinase that mediates its synthesis have not yet been determined. This pathway appears specific for ceramide derived from sphingomyelin, as ceramide derived from glycosphingolipids is not converted either to sphingoid bases (14) or to ceramide 1-phosphate (4). Recently, Hannun and coworkers (6-8) have provided evidence that this sphingomyelin pathway may be involved in signal transduction. Tumor necrosis factor (TNF) a, y interferon, and 1,25-dihydroxyvitamin D3, factors that induce monocytic differentiation of HL-60 promyelocytic cells, all stimulate sphingomyelin degradation to ceramide as an early
Leu-Val-Glu-Pro-Leu-Thr-Pro-Ser-Gly-Glu-Ala-Pro-AsnGln-Ala-Leu-Leu-Arg-COOH) was synthesized by using an Applied Biosystems model 431A machine and purified by reverse-phase HPLC. C8-ceramide (N-octanoylsphingosine; C8-cer) and TNF-a were gifts of R. Bittman (Queens College, New York) and D. Donner (Sloan-Kettering Institute), respectively. Cell Culture. A-431 human epidermoid carcinoma cells were grown in monolayer culture in a 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) and Ham's F-12 medium containing 10% FBS and were harvested by trypsinization as described (18). HL-60 cells were grown in suspension culture in RPMI 1640 medium containing 10% FBS and supplements, as described (3). On the day of an
The publication costs of this article were defrayed in part by page charge
Abbreviations: FBS, fetal bovine serum; TNF, tumor necrosis factor; C8-cer, C8-ceramide or N-octanoylsphingosine; EGFR, epidermal growth factor receptor.
payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
10009
10010
Cell
Biology: Mathias et al.
experiment HL-60 cells were resuspended (1 x 106 cells per ml) in RPMI 1640 medium/1% FBS for 2 hr before stimulation with lipid activators and differentiating agents. Membrane Preparation. Cells (3 X 107/ml) were homogenized with a tight-fitting Dounce homogenizer at 40C in buffer (25 mM Hepes, pH 7.4/5 mM EGTA/50 mM NaF/leupeptin at 10 Ag/ml) as described by Countaway et al. (17). The homogenate was centrifuged at 500 X g for 5 min, and the postnuclear supernatant was centrifuged at 200,000 x g for 30 min. The microsomal membrane pellet was resuspended (7.5 pug of membrane protein per ul for A-431 cells and 1.5 Ag/,l~ for HL-60 cells) in homogenizing buffer. Membranes were prepared fresh daily. Assay of Kinase Activity. For most experiments, the reaction mixture contained 25 Al of microsomal membrane or postnuclear supernatant, 50 ,ul of EGFR peptide (4 mg/ml in 25 mM Hepes, pH 7.4) and 125 A.l of buffer (50 mM Hepes, pH 7.4/20 mM MgCl2) (17). Phosphorylation was initiated at 22°C by addition of 50 pl of [yy-32P]ATP (150 ,M final concentration; 4000 dpm/pmol). For studies with lipid activators, ceramide and other lipids were dried under N2 and resuspended in the kinase assay buffer by bath sonication for 2 min at 37°C. The reaction was terminated at the indicated times by addition of 50 pu of 0.5 M ATP in 90% formic acid. Unless otherwise indicated, all assays were done under conditions determined as linear for time and enzyme concentration. Enzyme activity was determined from the transfer of 32p from the y position of ATP to EGFR peptide and the specific radioactivity of [y-32P]ATP. Phosphorylated peptide was quantified by two separate methods. For initial studies, samples were spotted on phosphocellulose paper, washed in 1 M acetic acid/4 mM pyrophosphate and subjected to liquid scintillation counting, as described (25). Values obtained from a boiled blank or a sample lacking peptide were subtracted from each determination. Alternatively, HPLC was done as described (17). For these studies, samples were first applied to a C18 Sep-Pak cartridge and eluted with 99.9% acetonitrile/0.1% trifluoroacetic acid. The eluates were lyophilized, resuspended in 6 M guanidine hydrochloride/200 mM Tris, pH 8.5 and applied to a C18 reverse-phase column (Dynamax, 4.6 mm i.d., Rainin, Woburn, MA). The peptide was eluted with a linear gradient (1% per min) of acetonitrile at a flow rate of 1 ml/min and was detected by measuring the Cerenkov radiation associated with 1-ml fractions. Phosphoamino Acid Analysis. To determine which amino acid was phosphorylated, phosphoamino acid analysis of the peptide was done. The phosphopeptide peak obtained by HPLC was subjected to partial acid hydrolysis (1 hr at 110°C in 6 M HCI). The hydrolysates were dried, resuspended in 250 ,ul of water, and applied to a Dowex AG1-X8 column (BioRad). Amino acids were eluted with 0.5 M HCI, dried, and analyzed by thin-layer electrophoresis, as described by Hunter and coworkers (26). Individual phosphoamino acids were identified by ninhydrin staining of carrier phosphoamino acids and by autoradiography. Other Procedures. Protein was measured by the method of Bradford (27). Statistics. Statistical analysis was performed by t test and linear regression analysis by the method of least squares.
RESULTS Davis and coworkers (15, 17) showed that addition of sphingosine to A-431 cells enhanced phosphorylation of the EGFR on Thr-669. Subsequently, work from this laboratory showed that sphingosine was rapidly converted to ceramide in these cells and that exogenous ceramide induced identical effects (18). To investigate the kinase that mediated ceramide action,
Proc. Natl. Acad. Sci. USA 88 (1991) a synthetic peptide corresponding to the sequence around Thr-669 was used as substrate. Initial studies were done to determine the kinetics of phosphorylation ofthe EGFR peptide. The conditions for this assay were adapted from Davis and coworkers (17, 25). Postnuclear supernatant was used as a source of enzyme activity, and samples were spotted on phosphocellulose paper to measure phosphorylated peptide. The kinetics of phosphorylation appeared biphasic. Initial rapid incorporation of 32p into peptide for 10 min was followed by incorporation at a slightly reduced rate for as long as 30 min (Fig. 1). Subsequent studies were done to optimize the assay. Kinase activity was found by Lineweaver-Burke analysis to be linearly related (r = 0.98) to substrate concentration for ATP (0.3-300 ,M) and EGFR peptide (0.02-1 mg/ml) at 5 min of stimulation. Apparent Km values for ATP of 15 ,uM and for EGFR peptide of 0.25 mg/ml were derived. Apparent Vmax values ranged from 100-200 pmol-min-1/mg of protein-'. All subsequent studies were done with 150 ,M ATP and EGFR peptide at 4 mg/ml. Under these conditions, substrate concentration was not rate limiting. An additional set of studies assessed the pH optimum for the kinase activity. There was no measurable activity at pH values
